MEASUREMENT OPERATION PARAMETER ADJUSTMENT APPARATUS, MACHINE LEARNING DEVICE, AND SYSTEM

• US 20190310115A1
• Filed: 04/04/2019
• Published: 10/10/2019
• Est. Priority Date: 04/05/2018
• Status: Active Grant

• Current Assignee: Fanuc Corporation
• Original Assignee: Fanuc Corporation
• Inventors: ABE, Hiroshi
• Granted Patent: US 11,579,000 B2
• Time in Patent Office: Days
• CPC Class Codes: G01D 18/00 Testing or calibrating apparatus or arrangements provided for in groups G01D1/00 - G01D15/00
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 19/18 Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form G05B19/418 takes precedence
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/08 Learning methods

Abstract

A measurement operation parameter adjustment apparatus that enables efficient measurement of the placement position of an object to be measured even in the case where there are variations in the placement positions, the sizes, and the product types of objects to be measured includes a machine learning device. The machine learning device observes measurement operation parameter data representing the measurement operation parameter of the measurement operation and measurement time data representing time taken to perform the measurement operation as a state variable representing a current environmental state and performs learning or decision-making using a learning model obtained by modeling adjustment of the measurement operation parameter based on the state variable.

Claims

1. A measurement operation parameter adjustment apparatus for adjusting a measurement operation parameter of a measurement operation that is executed by a measurement device for measuring a placement position of an object to be measured, the measurement operation parameter adjustment apparatus comprising:

a machine learning device for observing measurement operation parameter data representing the measurement operation parameter of the measurement operation and measurement time data representing time taken to perform the measurement operation as a state variable representing a current environmental state and performing learning or decision-making using a learning model obtained by modeling adjustment of the measurement operation parameter based on the state variable.

2. The measurement operation parameter adjustment apparatus according to claim 1, wherein

the machine learning device includes

a state observation section for observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state,

a determination data acquisition section for acquiring measurement operation determination data for determining whether the measurement operation executed based on the measurement operation parameter of the measurement operation is appropriate, as determination data representing an appropriate/inappropriate determination result for the measurement operation, and

a learning section for generating the learning model obtained by learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation using the state variable and the determination data.

3. The measurement operation parameter adjustment apparatus according to claim 2, wherein

the state observation section further observes product type data representing product type information on the object to be measured as the state variable representing the current environmental state, and

the learning section learns the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation and the product type information on the object to be measured.

4. The measurement operation parameter adjustment apparatus according to claim 2, wherein

the learning section includes

a reward calculation section for finding a reward relating to the appropriate/inappropriate determination result, and

a value function update section for, using the reward, updating a function representing a value of an action to adjust the measurement operation parameter of the measurement operation with respect to the time taken to perform the measurement operation, and

the reward calculation section gives higher rewards in cases where the time taken to perform the measurement operation is shorter and lower rewards in cases where the time taken to perform the measurement operation is longer, the object to be measured or the measurement device has been damaged, or the object to be measured has been moved.

5. The measurement operation parameter adjustment apparatus according to claim 1, wherein the learning section performs multi-layer calculation of the state variable and the determination data.

6. The measurement operation parameter adjustment apparatus according to claim 1, wherein

the machine learning device includes

a state observation section for observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state,

a learning section including the learning model obtained by learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation, and

a decision-making section for determining the adjustment of the measurement operation parameter of the measurement operation based on the state variable observed by the state observation section and the learning model.

7. The measurement operation parameter adjustment apparatus according to claim 1, wherein the machine learning device is located on a cloud server.

11. A system comprising:

a plurality of apparatuses connected to each other through a network, wherein

the plurality of apparatuses include at least a first measurement operation parameter adjustment apparatus according to claim 2.

12. The system according to claim 11, wherein

the plurality of apparatuses include a computer including a machine learning device,

the computer acquires at least one learning model generated by learning in the learning section of the first measurement operation parameter adjustment apparatus, and

the machine learning device in the computer performs optimization or streamlining based on the acquired learning model.

13. The system according to claim 11, wherein

the plurality of apparatuses include a second measurement operation parameter adjustment apparatus different from the first measurement operation parameter adjustment apparatus, and

a learning model generated by the learning section in the first measurement operation parameter adjustment apparatus is shared with the second measurement operation parameter adjustment apparatus.

14. The system according to claim 11, wherein

the plurality of apparatuses include a control device for controlling a measurement device, and

data regarding the measurement device, the data being acquired in the control device, are available for learning by the learning section in the first measurement operation parameter adjustment apparatus through the network.

8. A machine learning device for observing measurement operation parameter data representing a measurement operation parameter of a measurement operation that is executed by a measurement device for measuring a placement position of an object to be measured and measurement time data representing time taken to perform the measurement operation as a state variable representing a current environmental state, and performing learning or decision-making using a learning model obtained by modeling adjustment of the measurement operation parameter based on the state variable.

9. The machine learning device according to claim 8, comprising:

a state observation section for observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state;

a determination data acquisition section for acquiring measurement operation determination data for determining whether the measurement operation executed based on the measurement operation parameter of the measurement operation is appropriate, as determination data representing an appropriate/inappropriate determination result for the measurement operation; and

a learning section for generating the learning model obtained by learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation using the state variable and the determination data.

10. The machine learning device according to claim 8, comprising:

a state observation section for observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state;

a learning section including the learning model obtained by learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation; and

a decision-making section for determining the adjustment of the measurement operation parameter of the measurement operation based on the state variable observed by the state observation section and the learning model.

15. A method for performing learning or decision-making regarding adjustment of a measurement operation parameter of a measurement operation that is executed by a measurement device for measuring a placement position of an object to be measured, the method comprising:

a step of observing measurement operation parameter data representing the measurement operation parameter of the measurement operation and measurement time data representing time taken to perform the measurement operation as a state variable representing a current environmental state and performing learning or decision-making using a learning model obtained by modeling the adjustment of the measurement operation parameter based on the state variable.

16. A method regarding machine learning of adjustment of a measurement operation parameter of a measurement operation that is executed by a measurement device for measuring a placement position of an object to be measured, the method comprising:

a step of performing a process regarding the machine learning of the adjustment of the measurement operation parameter using a learning model obtained by modeling the adjustment of the measurement operation parameter based on measurement operation parameter data representing the measurement operation parameter of the measurement operation and measurement time data representing time taken to perform the measurement operation, the measurement operation parameter data and the measurement time data being observed as a state variable representing a current environmental state.

17. The method according to claim 16, wherein

the step of performing learning or decision-making using the learning model includes

a step of observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state,

a step of acquiring measurement operation determination data for determining whether the measurement operation executed based on the measurement operation parameter of the measurement operation is appropriate, as determination data representing an appropriate/inappropriate determination result for the measurement operation, and

a step of learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation using the state variable and the determination data.

18. The method according to claim 16, wherein

the step of performing learning or decision-making using the learning model includes

a step of observing the measurement operation parameter data representing the measurement operation parameter of the measurement operation and the measurement time data representing the time taken to perform the measurement operation as the state variable representing the current environmental state, and

a step of determining the adjustment of the measurement operation parameter of the measurement operation based on the observed state variable and a result of learning the adjustment of the measurement operation parameter of the measurement operation in relation to the time taken to perform the measurement operation.